Fourteen essays on how artificial intelligence is rewriting actuarial work: pricing, longevity, fraud, reserving, pensions, catastrophes, solvency, reinsurance, governance and the future of the profession. A new installment is published every three days.
Actuaries and data scientists share quantitative foundations, but the actuary brings risk judgment, the equivalence principle and professional accountability. A roadmap for converging without losing identity.
Read the articleThirteen more installments that work through the AI toolkit applied to risk, with actuarial judgment at every step.
Which investment problems machine learning truly solves for insurers and pension funds, where AI merely dresses up market noise, and why actuarial judgment governs the asset side of the balance sheet.
How machine learning and language models help read submissions, refine exposure rating and optimize reinsurance structures, without relieving the actuary of judgment on extreme losses.
Solvency regimes require nested simulations that take days. Proxy models and machine learning approximate the capital requirement in minutes, without displacing actuarial judgment on the tail.
How machine learning and climate models are transforming catastrophe modeling, from the event catalog to the exceedance curve, without relieving the actuary of judgment on the tail.
A tour of how property and casualty pricing has evolved, from the GLM as the standard to gradient boosting, without losing sight of interpretability and the actuarial framework.
From the classic Lee-Carter model to deep learning extensions: why neural networks are changing mortality projection and what interpretability challenges they raise for longevity, pensions and life insurance.
How machine learning detects anomalies, patterns and fraud rings, automates underwriting and speeds up scoring, and why explainability, bias and data quality remain the big challenges.
From the classic chain ladder to neural networks for loss triangles: how predictive analytics is reshaping reserve estimation and the quantification of uncertainty under Solvency II.
Insurance is adopting increasingly opaque models. Explainability, fairness and model risk governance define the frontier of responsible artificial intelligence, with the actuary accountable for the model.
What LLMs are, how they can support the actuary in documentation, code and analysis, and what safeguards and governance they require before you trust them.
How machine learning helps model longevity and pension system risk, how it relates to the projected unit credit method and NIF D-3, and the limits set by actuarial prudence.
How driving data and machine learning are transforming auto insurance pricing, from pay-how-you-drive to the challenges of privacy and fairness.
Why the combination of statistical rigor, business insight and professional standards makes AI an ally rather than a threat to the actuarial profession.
This is the publication order of the series. The first installment is already available; the rest are released on a staggered schedule.